rgapi 0.2.2__tar.gz → 0.2.4__tar.gz

{rgapi-0.2.2 → rgapi-0.2.4}/.github/workflows/ci.yml

@@ -62,7 +62,7 @@ jobs:
       - uses: dtolnay/rust-toolchain@stable
       - id: crates-auth
         uses: rust-lang/crates-io-auth-action@v1
-      - run: cargo publish -p rgapi
+      - run: cargo publish -p rgapi --no-verify
         env:
           CARGO_REGISTRY_TOKEN: ${{ steps.crates-auth.outputs.token }}
       - uses: actions/download-artifact@v8

{rgapi-0.2.2 → rgapi-0.2.4}/Cargo.lock

@@ -50,9 +50,9 @@ checksum = "7f8f80099a98041a3d1622845c271458a2d73e688351bf3cb999266764b81d48"
 
 [[package]]
 name = "crc32fast"
-version = "1.5.2"
+version = "1.5.4"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "01a7799fd6b852db0e61728dde9a204c423b44d689dbd432522543614b490e78"
+checksum = "5f97d3416d5337d15c4b0d78c8c91df922ca143f57ec20e2d1c188333047ef04"
 dependencies = [
  "cfg-if",
 ]
@@ -84,9 +84,9 @@ checksum = "a31eee39dddec8330830986fcd7625edb5a24ec90ea038215273bbc3adb08ac6"
 
 [[package]]
 name = "encoding_rs"
-version = "0.8.42"
+version = "0.8.44"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "8e985e0451871ad22fb8d2b6b076e2028a502a0d3950998c2c5c0a4f9b5d9679"
+checksum = "89253175e704ab869b04f9db224d38630e0211a53841a79181deae0889e169f5"
 dependencies = [
  "cfg-if",
  "core_detect",
@@ -107,9 +107,9 @@ dependencies = [
 
 [[package]]
 name = "fancy-regex"
-version = "0.19.2"
+version = "0.20.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "d301f5bf187b3c295fce6468d3875037a0bccc5f6b151c63cac2f85babf21912"
+checksum = "20ed7f62ce0ae0da1a8b93c3c7104ab7e0760a74491e097c701fff966e945390"
 dependencies = [
  "bit-set",
  "regex-automata",
@@ -118,9 +118,9 @@ dependencies = [
 
 [[package]]
 name = "globset"
-version = "0.4.20"
+version = "0.4.21"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "07c34a9410465b45bd9787443bc7370f37735bad04b0f0cd57ff1a3186c98988"
+checksum = "f98b0d85522dc0a824cfebbc9594ff4a9b289c568bea0370319dfbdf20ccdd2d"
 dependencies = [
  "aho-corasick",
  "bstr",
@@ -161,9 +161,9 @@ checksum = "2304e00983f87ffb38b55b444b5e3b60a884b5d30c0fca7d82fe33449bbe55ea"
 
 [[package]]
 name = "ignore"
-version = "0.4.33"
+version = "0.4.34"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "00b69833ed729dc5aa7d19541d96d6cf8e9137194207a04916d658e43168402f"
+checksum = "484be5657551d7868d67587df967fb4fcbe9d266a761cc855aaa148fda0b3cc7"
 dependencies = [
  "crossbeam-deque",
  "globset",
@@ -183,15 +183,15 @@ checksum = "8f42a60cbdf9a97f5d2305f08a87dc4e09308d1276d28c869c684d7777685682"
 
 [[package]]
 name = "libc"
-version = "0.2.190"
+version = "0.2.191"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "ce5d3ddc6d3fa000eb1536d85e147bfe31aacaba692ed6a876f95cb7c855be78"
+checksum = "65cd87c9a6d60536580b316ad985726104730a5151c99845f2490428ef91240a"
 
 [[package]]
 name = "log"
-version = "0.4.34"
+version = "0.4.35"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "f9f8bd3e56ce4dfc153cf470fffbfa98c7620958b312ca5c3a4b8d5181fd13c6"
+checksum = "5d15f86d12da44cd8d4ec3c73984537eb7f4cadcb9776a575a0a8a05ca11005c"
 
 [[package]]
 name = "memchr"
@@ -201,9 +201,9 @@ checksum = "cf8baf1c55e62ffcace7a9f06f4bd9cd3f0c4beb022d3b367256b91b87513d98"
 
 [[package]]
 name = "memmap2"
-version = "0.9.11"
+version = "0.9.12"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "d1219ed1b7f229ee7104d281dd01d6802fe28bb6e95d292942c4daacdeb798c0"
+checksum = "6b0ca1155a9ecc25aa12e8313a119f32edf0255fea2f44839c4ac0a12b2ac157"
 dependencies = [
  "libc",
 ]
@@ -237,9 +237,9 @@ dependencies = [
 
 [[package]]
 name = "pyo3"
-version = "0.29.3"
+version = "0.30.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "700d18fa267b73b9b521fd7e13580e2f446916f176cacee1ab63fcc8191f1655"
+checksum = "229ca13b8a15bb0d1487b81ab975bfcd97513416499a8312011b6036e42e0de4"
 dependencies = [
  "libc",
  "once_cell",
@@ -251,18 +251,18 @@ dependencies = [
 
 [[package]]
 name = "pyo3-build-config"
-version = "0.29.3"
+version = "0.30.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "7b3fc0c4d08f6bb10e71fe39dfb9e2f59c6eb6854e22ec8092f50c69a4499adb"
+checksum = "70130e2b6dd1ee589a8788026a429561ca5c40fa7a2fa7914c2282eddae24700"
 dependencies = [
  "target-lexicon",
 ]
 
 [[package]]
 name = "pyo3-ffi"
-version = "0.29.3"
+version = "0.30.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "dfc0b8e19df29aad7086cf977bb0c2a2f143e30567eb113e9cf72b62ca698330"
+checksum = "5e7ff73c555c286d212d4ce7e5fbda845a973c4f61bd5f160f5f2c3c2a04d1c4"
 dependencies = [
  "libc",
  "pyo3-build-config",
@@ -270,26 +270,26 @@ dependencies = [
 
 [[package]]
 name = "pyo3-macros"
-version = "0.29.3"
+version = "0.30.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "6100e8a4b5eba53afaa5ed078364851a0b2499c553a44c31026b929049b49dc6"
+checksum = "c46d2e5862c7bb8207e3b639f713dda3c499933444784154c11d013c5a691b60"
 dependencies = [
  "proc-macro2",
  "pyo3-macros-backend",
  "quote",
- "syn 2.0.119",
+ "syn",
 ]
 
 [[package]]
 name = "pyo3-macros-backend"
-version = "0.29.3"
+version = "0.30.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "6143877a16e82b5a727b7127ff4cd86858a24a28d745d72f43e6f227c7b1bdb3"
+checksum = "38b89f066f2ae72a2aef52d1537ca2fcd104440ebfadee06e97aa90738a25b9b"
 dependencies = [
  "heck",
  "proc-macro2",
  "quote",
- "syn 2.0.119",
+ "syn",
 ]
 
 [[package]]
@@ -320,7 +320,7 @@ checksum = "d6f6ff9a378485b298a5286656da665ba74413d36db0979633275d2e708145d4"
 
 [[package]]
 name = "rgapi"
-version = "0.2.2"
+version = "0.2.4"
 dependencies = [
  "crc32fast",
  "fancy-regex",
@@ -335,7 +335,7 @@ dependencies = [
 
 [[package]]
 name = "rgapi-py"
-version = "0.2.2"
+version = "0.2.4"
 dependencies = [
  "grep-matcher",
  "pyo3",
@@ -391,14 +391,14 @@ checksum = "e7a5d71263a5a7d47b41f6b3f06ba276f10cc18b0931f1799f710578e2309348"
 dependencies = [
  "proc-macro2",
  "quote",
- "syn 3.0.6",
+ "syn",
 ]
 
 [[package]]
 name = "serde_json"
-version = "1.0.151"
+version = "1.0.154"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "c841b55ecdae098c80dcae9cf767f6f8a0c2cdb3416bbef72181df4d0fe73f14"
+checksum = "e7e9cc8b1b85264074fbcc02a88680c4096b1e47df8f739dceb03bf482f04bd6"
 dependencies = [
  "itoa",
  "memchr",
@@ -415,20 +415,9 @@ checksum = "e3a9fe34e3e7a50316060351f37187a3f546bce95496156754b601a5fa71b76e"
 
 [[package]]
 name = "syn"
-version = "2.0.119"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "872831b642d1a07999a962a351ed35b955ea2cfc8f3862091e2a240a84f17297"
-dependencies = [
- "proc-macro2",
- "quote",
- "unicode-ident",
-]
-
-[[package]]
-name = "syn"
-version = "3.0.6"
+version = "3.0.9"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "8593e8e72159ed2257d083c7a454a85cbf854f37a0966d8d483aff8c8a3ebcee"
+checksum = "d78c8dee4c7bf0e14673097256fed6142ce9d3b85a408189d07482442145823b"
 dependencies = [
  "proc-macro2",
  "quote",
@@ -443,9 +432,9 @@ checksum = "adb6935a6f5c20170eeceb1a3835a49e12e19d792f6dd344ccc76a985ca5a6ca"
 
 [[package]]
 name = "unicode-ident"
-version = "1.0.26"
+version = "1.0.27"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "d245f478577f809a851594d02313b640fb437e0bb33866753cff937863096954"
+checksum = "a2c754d6c33795a1c324727428e5a7dedb5b06195f9890bdbcba760d3e246563"
 
 [[package]]
 name = "walkdir"

{rgapi-0.2.2 → rgapi-0.2.4}/Cargo.toml

@@ -2,7 +2,7 @@
 members = ["py"]
 
 [workspace.package]
-version = "0.2.2"
+version = "0.2.4"
 edition = "2024"
 
 [workspace.dependencies]
@@ -24,7 +24,7 @@ name = "rgapi"
 
 [dependencies]
 crc32fast = "1"
-fancy-regex = "0.19"
+fancy-regex = ">=0.20"
 globset = ">=0.4.18"
 grep-matcher.workspace = true
 grep-searcher = ">=0.1.16"

{rgapi-0.2.2 → rgapi-0.2.4}/DEV.md

@@ -23,7 +23,7 @@ cargo develop
 pytest -q
 ```
 
-`cargo develop` and bare `cargo test` use the same Cargo profile and share the core and binding libraries. Unit tests also build a separate `cfg(test)` executable. The published `rgapi` crate has no Python dependency; the unpublished `rgapi-py` crate in `py/` builds the extension. Run `cargo fmt` after Rust edits and `chkstyle` before committing Python edits.
+`cargo develop` and bare `cargo test` use the same Cargo profile and share the core and binding libraries. Unit tests also build a separate `cfg(test)` executable. The published `rgapi` crate has no Python dependency; the unpublished `rgapi-py` crate in `py/` builds the extension. Run `cargo fastfmt` after Rust edits and `chkstyle` before committing Python edits.
 
 ## Release
 
@@ -41,7 +41,9 @@ The GitHub workflow builds wheels for Python 3.10-3.13 on Linux and macOS and pu
 
 All user-supplied regexes use the shared `RegexMatcher` in `src/matcher.rs`, backed by `fancy-regex`. There is no separate fallback engine or native C dependency. The adapter implements `grep_matcher::Matcher` so `grep-searcher` continues to handle file scanning, line numbers, context, and binary detection. `compile_regex` returns rgapi's own exported `RegexMatcher` type.
 
-The matcher centralizes smart-case analysis and line-ending policy. Line-safe patterns can search a whole buffer; patterns that can consume line endings or inspect neighbouring lines are evaluated on individual lines. Notebook multiline matching instead uses the whole cell. Filename/path regexes match their whole input. Backtracking-limit errors propagate through the existing result channels, including directory-pruning errors; cancellation must not hide a queued error from async consumers.
+The matcher centralizes smart-case analysis and line-ending policy. Line-safe patterns can search a whole buffer; patterns that can consume line endings, inspect neighbouring lines or need the backtracking engine are evaluated on individual lines, so that engine's step budget covers one line, not a whole buffer. Notebook multiline matching instead uses the whole cell. Filename/path regexes match their whole input. Backtracking-limit errors propagate through the existing result channels, including directory-pruning errors; cancellation must not hide a queued error from async consumers.
+
+A walk error, a file the walk can't read, or a notebook `nbrg` can't parse is a skipped error (`RgApiError::is_skipped`). `send_error` sends a skipped error without ending the walk. Any other error ends it. The Python bindings issue a `RuntimeWarning` for each skipped error and raise `ValueError` for any other (`skip_or_raise`, and `worker_error` on async worker threads). Rust's collecting functions, such as `find`, `rg`, `nb_search` and `StreamIter::collect_timeout`, return the first error of either kind. `entry_err` drops a permission error silently at the `max_depth` cap, where the walker opens a directory it never descends into.
 
 Python discovery and `paths=True` results contain absolute `pathlib.Path` objects. Structured search rows hold base-relative string labels with `/` separators. Traversal uses `ignore::WalkParallel`, so result order is not part of the API contract. Search results are structured rows; collected result lists use rg-style `str()` and notebook display. `SearchLine.lnhash` is computed with the same CRC-32-based line-content hash format as exhash (`lineno|hash|`, low 12 bits of CRC-32 over the line's UTF-8 bytes, encoded as two Base64url characters); `lnhashs=True` only changes row display, not `line_number` or matching behavior. Path regexes filter returned/searched paths; `skip_dir` and `skip_dir_re` prune traversal through `ignore::WalkBuilder::filter_entry`. Depth, size, filesystem, hidden, and ignore options use `ignore::WalkBuilder` settings. The `ignore` walker follows every root it is given. Discovery checks each root with `symlink_metadata`. The worker sends each unfollowed link root as a result before walking the other roots. This also supports dangling roots, which the underlying walker would reject. `ls` sets `walk_root_links`, which gives a root link to a directory to the walker. The walker then reports paths under the link. Other discovery roots use absolute paths without canonicalizing. Content searches retain canonical root resolution. `rg_iter` exposes the same parallel search stream that `rg` collects by default; `paths=True` and `count=True` consume that stream with different reducers. Text search skips binary files and invalid UTF-8 content.
 
@@ -51,9 +53,9 @@ Streaming engine: `walk.rs` owns the generic machinery. `StreamIter<T>` is the w
 
 Async API: `fda`, `fda_iter`, `rga`, `rga_iter`, `nbrga`, and `nbrga_iter` wrap the corresponding private core operations. `rga(summary=True)` uses `_core.block_search_async`; ordinary `rga` uses `_core.rg_async`. Each collected core function takes a Python callback, runs on Rust threads through the generic `stream_async` helper, and delivers with one GIL attach at the end. Iterator forms use `stream_iter_async` and attach once per batch. The Python side settles an `asyncio.Future` or feeds an `asyncio.Queue` via `loop.call_soon_threadsafe`; no Python thread blocks and `asyncio.to_thread` is not involved. `AsyncHandle.cancel()` sets the same atomic flag used by the Rust iterators.
 
-Truncation is recorded on collected results: `max_results` sets `stop_reason="max_results"`, and `timeout_ms` on `rg`/`rga`/`nbrg`/`nbrga`/`fd`/`fda`/`walk`/`ls` sets `stop_reason="timeout"`. `SearchResults`, `BlockResults`, `PathResults`, and `NbResults` share this through `_Results`; `complete` means `stop_reason is None`. In block summary mode, `max_results` counts matching blocks and keeps their block context. `count=True` returns a plain int, so it rejects timeouts and block summary mode.
+Truncation is recorded on collected results: `max_results` sets `stop_reason="max_results"`, and `timeout_ms` on `rg`/`rga`/`nbrg`/`nbrga`/`fd`/`fda`/`ls` sets `stop_reason="timeout"`. `SearchResults`, `BlockResults`, `PathResults`, and `NbResults` share this through `_Results`; `complete` means `stop_reason is None`. In block summary mode, `max_results` counts matching blocks and keeps their block context. `count` returns a plain int, or a dict for `count="file"`, so it rejects timeouts and block summary mode.
 
-Rust discovery returns base-relative `PathBuf` values. Glob filters operate on native paths; only regex matching uses string labels. PyO3 converts discovery roots and results with its filesystem-path support. Python joins each result to the base from `_core.walk_base` without resolving links. `PathResults` stores that display base and completion status, including across slices. Its `__repr__` uses `Path.lstat()` for an `ls -l`-style listing capped at `MAX_REPR` rows. `str()` returns one base-relative name per line. `ls` is `fd` with shell-style defaults (one level, dirs, ignore rules off) and `walk_root_links`, sorted in place.
+Rust discovery returns base-relative `PathBuf` values. Glob filters operate on native paths; only regex matching uses string labels. PyO3 converts discovery roots and results with its filesystem-path support. Python joins each result to the base from `_core.walk_base` without resolving links. `PathResults` stores that display base and completion status, including across slices. Its `__repr__` uses `Path.lstat()` for an `ls -l`-style listing capped at `MAX_REPR` rows. `str()` returns one base-relative name per line. `ls` is `fd` with shell-style defaults (one level, dirs, ignore rules off) and `walk_root_links`, sorted in place. `fd` takes `pattern` first, as `rg` does. `ls` takes `root` first, as the shell's `ls DIR` does.
 
 Rust callers can consume a `StreamIter` with `cancel_and_join()` to cancel, drain queued results, and wait for the walk's workers to finish. `Drop` remains nonblocking. Since filesystem calls already in progress must return before joining completes, keep `cancel_and_join()` off async executors.
 

{rgapi-0.2.2 → rgapi-0.2.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: rgapi
-Version: 0.2.2
+Version: 0.2.4
 Classifier: Programming Language :: Rust
 Classifier: Programming Language :: Python :: Implementation :: CPython
 Requires-Dist: fastcore>=2.2.29
@@ -32,7 +32,7 @@ For common file discovery and search:
 ```python
 from rgapi import fd, ls, rg, rg_iter
 
-fd(".", ext="py", exclude="test_*.py")
+fd(ext="py", exclude="test_*.py")
 ls("src")
 for row in rg_iter("TODO", ".", include="*.py", context=2): print(row.asdict())
 rg("TODO", ".", ext="py", skip_dir=".venv", paths=True)
@@ -56,16 +56,15 @@ await rga("TODO", ".", ext="py", timeout_ms=200)
 async for row in rga_iter("TODO", "."): print(row)
 ```
 
-Use the lower-level functions to compile a regex, search text or a single file, or walk a directory:
+Use the lower-level functions to compile a regex, or to search text or a single file:
 
 ```python
-from rgapi import compile, search_path, search_text, walk
+from rgapi import compile, search_path, search_text
 
 matcher = compile("TODO")
 matcher.is_match("TODO")
 matcher.finditer("TODO TODO")
 
-walk(".")
 search_text(matcher, "alpha\nTODO\nomega\n", path="memory.txt", context=1)
 search_path(matcher, "src/lib.rs", display_path="src/lib.rs")
 ```
@@ -78,7 +77,7 @@ pip install rgapi
 
 ## File discovery
 
-`fd` and `walk` return absolute `pathlib.Path` objects. Use them directly with `.read_text()`, `.open()`, `.stat()`, or other filesystem operations. Their collected results display names relative to `root`. Pass `root` as a `str` or `Path`, or as a list of them. The sync and async APIs expand `~` and accept `.`, `./`, and paths containing `..`.
+`fd` returns absolute `pathlib.Path` objects. Use them directly with `.read_text()`, `.open()`, `.stat()`, or other filesystem operations. Its collected results display names relative to `root`. Pass `root` as a `str` or `Path`, or as a list of them. The sync and async APIs expand `~` and accept `.`, `./`, and paths containing `..`.
 
 A list of roots, such as `rg("TODO", ["src", "tests"])`, works with every walk and search function. The roots are searched as one walk that returns one result list. `timeout_ms` and `max_results` apply to the whole walk. Result paths are relative to the common ancestor of the roots. For example, rows read `src/app.py` and `tests/test_app.py`. `PathResults` displays use the same relative paths. Filters such as `include`, `path_re` and `skip_dir` match these relative paths. A file under more than one root, such as `src/app.py` with roots `[".", "src"]`, appears once.
 
@@ -86,9 +85,11 @@ Discovery uses the `ignore` crate with ripgrep's default filters. It reads `.git
 
 Hidden files are skipped unless `hidden=True`. Discovery returns symlinks themselves, including explicitly named roots and dangling links. Set `follow_links=True` to traverse their targets. Paths retain the symlink spelling. Use `same_file_system=True` to avoid crossing filesystem boundaries.
 
+A walk skips each directory or file that it can't read, such as one without read permission, and issues a `RuntimeWarning` that names it. The rest of the walk continues, as ripgrep's does.
+
 Traversal runs in parallel without guaranteed result order. Use `sorted(...)` when order matters.
 
-`fd` adds filename filters to `walk`. Its `pattern` is a smart-case regex matched against each basename. Lowercase patterns match case-insensitively. A pattern containing uppercase letters is case-sensitive. Use `path_re` to match the slash-separated relative path instead.
+`fd` takes `pattern` first and `root` second, as `rg` does. With no pattern, it returns every file under `root`. `pattern` is a smart-case regex matched against each basename. Lowercase patterns match case-insensitively. A pattern containing uppercase letters is case-sensitive. Use `path_re` to match the slash-separated relative path instead.
 
 `include` and `exclude` use case-sensitive glob syntax. `glob=` is an alias for `include=`. Patterns without `/` match names at any depth, so `*.py` matches `src/app.py`. Patterns containing `/` are root-relative. `*` stays within a path component; `**` spans directories: `src/*` matches immediate children, whereas `src/**` also matches deeper descendants. Matching a directory with `exclude` prunes its entire subtree. Includes never prune traversal. Excludes always win, and includes do not override ignore rules.
 
@@ -96,7 +97,7 @@ Filter extensions with `ext="py"` or `ext=["py", "rs"]`. Extension and glob filt
 
 Set `min_depth` and `max_depth` to bound recursion. `max_filesize` skips files above a byte limit.
 
-`ls` follows the shell command's listing conventions. It uses `fd` with `max_depth=1`, includes directories, disables ignore rules, and sorts by name. Set `hidden=True` for `ls -a` behaviour. All `fd` filters remain available.
+`ls` follows the shell command's listing conventions. It takes `root` first, as the shell's `ls DIR` does. It uses `fd` with `max_depth=1`, includes directories, disables ignore rules, and sorts by name. Set `hidden=True` for `ls -a` behaviour. All `fd` filters remain available.
 
 `fd_iter` yields absolute `Path` objects as the walk finds them. It accepts every `fd` filter. Stopping iteration ends the walk. It does not accept `timeout_ms`.
 
@@ -113,10 +114,10 @@ Patterns support backreferences and lookaround everywhere, including filename an
 ```python
 rg(r"\b(\w+)\s+\1\b", ".")  # repeated words
 rg(r"(?<=TODO: )\w+", ".")   # word following TODO:
-fd(".", pattern=r"^(\w+)_\1\.py$")  # names such as utils_utils.py
+fd(r"^(\w+)_\1\.py$")  # names such as utils_utils.py
 ```
 
-All regex parameters use `fancy-regex`; no engine-selection flag or C library is required. Ordinary patterns use its non-backtracking engine. Patterns requiring backtracking have a bounded backtracking budget; exceeding it raises `ValueError`, rather than silently skipping a file or path.
+All regex parameters use `fancy-regex`, with no engine-selection flag and no C library. Ordinary patterns, including word boundaries (`\b`, `\B`) outside lookbehinds, use its non-backtracking engine. Patterns requiring backtracking have a step budget that allows lines of millions of characters. A search that exceeds the budget raises `ValueError`. It never silently skips a file or path.
 
 Each `SearchLine` has these fields:
 
@@ -135,7 +136,7 @@ matches      list of (start, end) byte offsets for match rows
 
 Pass `lnhashs=True` to `rg` or `rg_iter` to display hash addresses instead of line numbers. The `line_number` field remains available.
 
-For other result forms, use `rg(..., paths=True)` to return unique matched paths or `rg(..., count=True)` to count match spans. `paths` and `count` cannot both be set.
+For other result forms, use `rg(..., paths=True)` to return unique matched paths, `rg(..., count=True)` to count match spans, or `rg(..., count="file")` for a dict from each file's label to its count, sorted by label. `paths` and `count` cannot both be set.
 
 `^` and `$` match at the start and end of each line. A line ends at `\n` or `\r\n`. `$` also matches before a `\r` that has no `\n` after it. `rg`, `rgstr`, and `compile` reject patterns containing a literal `\n` or `\r`. `nbrg` allows them with `multiline=True`.
 
@@ -143,10 +144,10 @@ Content searches and compiled content matchers are confined to individual lines:
 
 ### Path results
 
-`fd`, `walk`, and `ls` return `PathResults`, a list of absolute `Path` objects. So do `rg` and `nbrg` with `paths=True`, and their async equivalents. Indexing or iterating returns ordinary Paths:
+`fd` and `ls` return `PathResults`, a list of absolute `Path` objects. So do `rg` and `nbrg` with `paths=True`, and their async equivalents. Indexing or iterating returns ordinary Paths:
 
 ```python
-for path in fd("src", ext="py"):
+for path in fd(root="src", ext="py"):
     text = path.read_text()
     size = path.stat().st_size
 ```
@@ -163,13 +164,13 @@ Rust callers can set `FindOptions::special_files` to also discover FIFOs, socket
 
 ### Limits and timeouts
 
-Set `timeout_ms` on `rg`, `fd`, `walk`, or `ls` to stop at a deadline and return the results collected so far. Their async versions accept it too. Results report why the operation stopped:
+Set `timeout_ms` on `rg`, `fd` or `ls` to stop at a deadline and return the results collected so far. Their async versions accept it too. Results report why the operation stopped:
 
 - `stop_reason=None` means the result is complete.
 - `stop_reason="max_results"` means `max_results` truncated the result.
 - `stop_reason="timeout"` means the deadline was reached.
 
-`complete` is true exactly when `stop_reason` is `None`. `count=True` returns a plain integer without a completion flag. It cannot be combined with `timeout_ms`.
+`complete` is true exactly when `stop_reason` is `None`. A count is a plain integer or dict, with no completion flag. Counting can't be combined with `timeout_ms`.
 
 Deadlines stop result collection and request worker cancellation. Workers cannot interrupt a regex evaluation or filesystem call already in progress; a timeout is not a hard per-match execution limit.
 
@@ -236,7 +237,7 @@ A matching cell's display starts at its first matched line. Earlier lines are re
 - `count=True` returns the number of matching cells.
 - `timeout_ms` applies a deadline with the same `stop_reason` values as `rg`.
 
-The parser reads only each cell's `id`, `cell_type`, and `source`. It skips outputs and metadata without loading embedded images or plots. `search_nb(pattern, path, ...)` searches a single notebook file in the same way.
+The parser reads only each cell's `id`, `cell_type`, and `source`. It skips outputs and metadata without loading embedded images or plots. `search_nb(pattern, path, ...)` searches a single notebook file in the same way. `nbrg` skips a `.ipynb` file that isn't valid JSON and issues a `RuntimeWarning` naming the file and the parse error. `search_nb` raises `ValueError` for such a file.
 
 `rgapi-nbrg` exposes notebook search without requiring a Python kernel:
 
@@ -256,7 +257,7 @@ Run `rgapi-nbrg --help` for its discovery, matching, and output options.
 ```python
 from rgapi import fda, rga, rga_iter
 
-await fda(".", ext="py")
+await fda(ext="py")
 res = await rga("TODO", ".", timeout_ms=200)
 if not res.complete: print(f"partial results: {res.stop_reason}")
 async for row in rga_iter("TODO", "."): ...

{rgapi-0.2.2 → rgapi-0.2.4}/README.md

@@ -11,7 +11,7 @@ For common file discovery and search:
 ```python
 from rgapi import fd, ls, rg, rg_iter
 
-fd(".", ext="py", exclude="test_*.py")
+fd(ext="py", exclude="test_*.py")
 ls("src")
 for row in rg_iter("TODO", ".", include="*.py", context=2): print(row.asdict())
 rg("TODO", ".", ext="py", skip_dir=".venv", paths=True)
@@ -35,16 +35,15 @@ await rga("TODO", ".", ext="py", timeout_ms=200)
 async for row in rga_iter("TODO", "."): print(row)
 ```
 
-Use the lower-level functions to compile a regex, search text or a single file, or walk a directory:
+Use the lower-level functions to compile a regex, or to search text or a single file:
 
 ```python
-from rgapi import compile, search_path, search_text, walk
+from rgapi import compile, search_path, search_text
 
 matcher = compile("TODO")
 matcher.is_match("TODO")
 matcher.finditer("TODO TODO")
 
-walk(".")
 search_text(matcher, "alpha\nTODO\nomega\n", path="memory.txt", context=1)
 search_path(matcher, "src/lib.rs", display_path="src/lib.rs")
 ```
@@ -57,7 +56,7 @@ pip install rgapi
 
 ## File discovery
 
-`fd` and `walk` return absolute `pathlib.Path` objects. Use them directly with `.read_text()`, `.open()`, `.stat()`, or other filesystem operations. Their collected results display names relative to `root`. Pass `root` as a `str` or `Path`, or as a list of them. The sync and async APIs expand `~` and accept `.`, `./`, and paths containing `..`.
+`fd` returns absolute `pathlib.Path` objects. Use them directly with `.read_text()`, `.open()`, `.stat()`, or other filesystem operations. Its collected results display names relative to `root`. Pass `root` as a `str` or `Path`, or as a list of them. The sync and async APIs expand `~` and accept `.`, `./`, and paths containing `..`.
 
 A list of roots, such as `rg("TODO", ["src", "tests"])`, works with every walk and search function. The roots are searched as one walk that returns one result list. `timeout_ms` and `max_results` apply to the whole walk. Result paths are relative to the common ancestor of the roots. For example, rows read `src/app.py` and `tests/test_app.py`. `PathResults` displays use the same relative paths. Filters such as `include`, `path_re` and `skip_dir` match these relative paths. A file under more than one root, such as `src/app.py` with roots `[".", "src"]`, appears once.
 
@@ -65,9 +64,11 @@ Discovery uses the `ignore` crate with ripgrep's default filters. It reads `.git
 
 Hidden files are skipped unless `hidden=True`. Discovery returns symlinks themselves, including explicitly named roots and dangling links. Set `follow_links=True` to traverse their targets. Paths retain the symlink spelling. Use `same_file_system=True` to avoid crossing filesystem boundaries.
 
+A walk skips each directory or file that it can't read, such as one without read permission, and issues a `RuntimeWarning` that names it. The rest of the walk continues, as ripgrep's does.
+
 Traversal runs in parallel without guaranteed result order. Use `sorted(...)` when order matters.
 
-`fd` adds filename filters to `walk`. Its `pattern` is a smart-case regex matched against each basename. Lowercase patterns match case-insensitively. A pattern containing uppercase letters is case-sensitive. Use `path_re` to match the slash-separated relative path instead.
+`fd` takes `pattern` first and `root` second, as `rg` does. With no pattern, it returns every file under `root`. `pattern` is a smart-case regex matched against each basename. Lowercase patterns match case-insensitively. A pattern containing uppercase letters is case-sensitive. Use `path_re` to match the slash-separated relative path instead.
 
 `include` and `exclude` use case-sensitive glob syntax. `glob=` is an alias for `include=`. Patterns without `/` match names at any depth, so `*.py` matches `src/app.py`. Patterns containing `/` are root-relative. `*` stays within a path component; `**` spans directories: `src/*` matches immediate children, whereas `src/**` also matches deeper descendants. Matching a directory with `exclude` prunes its entire subtree. Includes never prune traversal. Excludes always win, and includes do not override ignore rules.
 
@@ -75,7 +76,7 @@ Filter extensions with `ext="py"` or `ext=["py", "rs"]`. Extension and glob filt
 
 Set `min_depth` and `max_depth` to bound recursion. `max_filesize` skips files above a byte limit.
 
-`ls` follows the shell command's listing conventions. It uses `fd` with `max_depth=1`, includes directories, disables ignore rules, and sorts by name. Set `hidden=True` for `ls -a` behaviour. All `fd` filters remain available.
+`ls` follows the shell command's listing conventions. It takes `root` first, as the shell's `ls DIR` does. It uses `fd` with `max_depth=1`, includes directories, disables ignore rules, and sorts by name. Set `hidden=True` for `ls -a` behaviour. All `fd` filters remain available.
 
 `fd_iter` yields absolute `Path` objects as the walk finds them. It accepts every `fd` filter. Stopping iteration ends the walk. It does not accept `timeout_ms`.
 
@@ -92,10 +93,10 @@ Patterns support backreferences and lookaround everywhere, including filename an
 ```python
 rg(r"\b(\w+)\s+\1\b", ".")  # repeated words
 rg(r"(?<=TODO: )\w+", ".")   # word following TODO:
-fd(".", pattern=r"^(\w+)_\1\.py$")  # names such as utils_utils.py
+fd(r"^(\w+)_\1\.py$")  # names such as utils_utils.py
 ```
 
-All regex parameters use `fancy-regex`; no engine-selection flag or C library is required. Ordinary patterns use its non-backtracking engine. Patterns requiring backtracking have a bounded backtracking budget; exceeding it raises `ValueError`, rather than silently skipping a file or path.
+All regex parameters use `fancy-regex`, with no engine-selection flag and no C library. Ordinary patterns, including word boundaries (`\b`, `\B`) outside lookbehinds, use its non-backtracking engine. Patterns requiring backtracking have a step budget that allows lines of millions of characters. A search that exceeds the budget raises `ValueError`. It never silently skips a file or path.
 
 Each `SearchLine` has these fields:
 
@@ -114,7 +115,7 @@ matches      list of (start, end) byte offsets for match rows
 
 Pass `lnhashs=True` to `rg` or `rg_iter` to display hash addresses instead of line numbers. The `line_number` field remains available.
 
-For other result forms, use `rg(..., paths=True)` to return unique matched paths or `rg(..., count=True)` to count match spans. `paths` and `count` cannot both be set.
+For other result forms, use `rg(..., paths=True)` to return unique matched paths, `rg(..., count=True)` to count match spans, or `rg(..., count="file")` for a dict from each file's label to its count, sorted by label. `paths` and `count` cannot both be set.
 
 `^` and `$` match at the start and end of each line. A line ends at `\n` or `\r\n`. `$` also matches before a `\r` that has no `\n` after it. `rg`, `rgstr`, and `compile` reject patterns containing a literal `\n` or `\r`. `nbrg` allows them with `multiline=True`.
 
@@ -122,10 +123,10 @@ Content searches and compiled content matchers are confined to individual lines:
 
 ### Path results
 
-`fd`, `walk`, and `ls` return `PathResults`, a list of absolute `Path` objects. So do `rg` and `nbrg` with `paths=True`, and their async equivalents. Indexing or iterating returns ordinary Paths:
+`fd` and `ls` return `PathResults`, a list of absolute `Path` objects. So do `rg` and `nbrg` with `paths=True`, and their async equivalents. Indexing or iterating returns ordinary Paths:
 
 ```python
-for path in fd("src", ext="py"):
+for path in fd(root="src", ext="py"):
     text = path.read_text()
     size = path.stat().st_size
 ```
@@ -142,13 +143,13 @@ Rust callers can set `FindOptions::special_files` to also discover FIFOs, socket
 
 ### Limits and timeouts
 
-Set `timeout_ms` on `rg`, `fd`, `walk`, or `ls` to stop at a deadline and return the results collected so far. Their async versions accept it too. Results report why the operation stopped:
+Set `timeout_ms` on `rg`, `fd` or `ls` to stop at a deadline and return the results collected so far. Their async versions accept it too. Results report why the operation stopped:
 
 - `stop_reason=None` means the result is complete.
 - `stop_reason="max_results"` means `max_results` truncated the result.
 - `stop_reason="timeout"` means the deadline was reached.
 
-`complete` is true exactly when `stop_reason` is `None`. `count=True` returns a plain integer without a completion flag. It cannot be combined with `timeout_ms`.
+`complete` is true exactly when `stop_reason` is `None`. A count is a plain integer or dict, with no completion flag. Counting can't be combined with `timeout_ms`.
 
 Deadlines stop result collection and request worker cancellation. Workers cannot interrupt a regex evaluation or filesystem call already in progress; a timeout is not a hard per-match execution limit.
 
@@ -215,7 +216,7 @@ A matching cell's display starts at its first matched line. Earlier lines are re
 - `count=True` returns the number of matching cells.
 - `timeout_ms` applies a deadline with the same `stop_reason` values as `rg`.
 
-The parser reads only each cell's `id`, `cell_type`, and `source`. It skips outputs and metadata without loading embedded images or plots. `search_nb(pattern, path, ...)` searches a single notebook file in the same way.
+The parser reads only each cell's `id`, `cell_type`, and `source`. It skips outputs and metadata without loading embedded images or plots. `search_nb(pattern, path, ...)` searches a single notebook file in the same way. `nbrg` skips a `.ipynb` file that isn't valid JSON and issues a `RuntimeWarning` naming the file and the parse error. `search_nb` raises `ValueError` for such a file.
 
 `rgapi-nbrg` exposes notebook search without requiring a Python kernel:
 
@@ -235,7 +236,7 @@ Run `rgapi-nbrg --help` for its discovery, matching, and output options.
 ```python
 from rgapi import fda, rga, rga_iter
 
-await fda(".", ext="py")
+await fda(ext="py")
 res = await rga("TODO", ".", timeout_ms=200)
 if not res.complete: print(f"partial results: {res.stop_reason}")
 async for row in rga_iter("TODO", "."): ...

{rgapi-0.2.2 → rgapi-0.2.4}/py/Cargo.toml

@@ -11,7 +11,7 @@ test = false
 [dependencies]
 rgapi = { path = ".." }
 grep-matcher.workspace = true
-pyo3 = { version = ">=0.29", features = ["extension-module"] }
+pyo3 = { version = ">=0.30", features = ["extension-module"] }
 
 [build-dependencies]
-pyo3-build-config = "0.29"
+pyo3-build-config = ">=0.29"